PROJECT SUMMARY/ABSTRACT Alzheimer's disease (AD) causes altered neuronal network activity and cognitive decline, but the underlying cellular and circuit mechanisms of network dysfunction remain unclear. The brain relies on oscillatory rhythms generated by inhibitory interneurons to precisely time neuronal firing required for circuit functions. We, and others, have postulated that impaired inhibitory function is a key mechanism of network hyperactivity and dysrhythmia in AD and related disease models. However, previous studies of this hypothesis used methods that lack the cellular resolution to interrogate cell and circuit mechanisms of network dysfunction. We propose to test the overarching hypothesis that network dysfunction (e.g., dysrhythmia) results from altered cellular (e.g., firing rates) and circuit (e.g., disinhibition) mechanisms of specific neuronal cell types. The two major interneuron types, fast-spiking parvalbumin (PV) and non-fast-spiking somatostatin (SOM) interneurons, regulate circuit functions by respectively promoting high-frequency gamma oscillations or slow- frequency beta oscillations. PV cell-dependent gamma oscillations increase within active networks during memory encoding and predict memory formation in humans and mice, whereas SOM-cell-dependent beta oscillations characterize resting networks. Remarkably, brain rhythms and synaptic activity seem to regulate A? deposition in humans and mice, but a cohesive hypothesis at the cellular and circuit level has not been proposed or tested. We hypothesize that network dysfunction results from hypoactive PV cells and hyperactive SOM cells, which reduce the pro-cognitive and ?pro-A?-clearance? PV cell-driven gamma oscillations and increases the pro-resting and ?pro-A?-accumulation? SOM cell-driven beta oscillations. We will use state-of-the-art in vivo electrophysiology to simultaneously record local field potentials (LFP) and multi single-unit activity of distinct inhibitory cell types and principal cells (PV, SOM, and PC cells) at baseline or during optogenetic stimulation of specific interneuron cell types to identify cell (e.g., reduced firing rate, impaired action potential waveform), circuit (e.g., reduced feed-forward inhibition), or network (e.g., altered gamma oscillations) mechanism of network dysfunction in the J20 and APP-KI mouse models. We specifically prose to: Aim 1, Determine the interneuron cell types that contribute to normal or altered gamma oscillations in non- transgenic (NTG) and J20 mice (Aim 1a) and that modulate A? levels in J20 mice (Aim 1b) in vivo; Aim 2, Determine whether altered single-unit activity (action potential) of interneuron cell types and principal cells are linked to network oscillatory deficits at baseline and during ChR2-evoked gamma oscillations in J20 and APP-KI mice in vivo; and Aim 3, Determine if optogenetic activation of transplanted PV cells or silencing endogenous SOM cells restores cell type-specific oscillations, circuit functions, and reduce A? levels in J20 and APP-KI mice.